This site is a work in progress. New lessons are added regularly. Contact us

Arithmetic of Complex Numbers

Learning goals

  • Derive every operation from i2=1i^2 = -1 plus ordinary algebra
  • Add and subtract part by part
  • Expand a product, then exchange i2i^2 for 1-1
  • Use zz=a2+b2z\overline{z} = a^2 + b^2, real and never negative
  • Show conjugation passing through sums and products
  • Divide by rationalizing with the denominator's conjugate

One new symbol, no new laws

When the previous lesson called a+bia + bi a number, that word carried a promise: the new numbers obey the same laws of algebra as the old ones. Addition is still commutative and associative, multiplication still distributes over addition, and subtraction and division still mean what they always meant, namely adding an opposite and multiplying by a reciprocal. The complex numbers add exactly one fact to that rulebook:

i2=1.i^2 = -1.

Everything in this lesson is forced by that single stipulation. The working strategy throughout is to treat a+bia + bi as a two-term expression in the symbol ii and run the ordinary algebra you have used for years. Then exchange i2i^2 for 1-1 the moment it appears. The question worth asking in advance is whether the answers stay inside the system. If you add, subtract, multiply, or divide two complex numbers, is the result again of the form (real)+(real)i(\text{real}) + (\text{real})\,i? Watch for the answer as each operation is built, because it will be yes all four times. That is what makes the complex numbers a self-contained place to do mathematics rather than a notational accident.

Before computing anything, pin down what it means for two complex answers to agree. Your work will produce expressions like 23+2i23 + 2i, and you need to know that no other pair of real numbers can name that same value.

Equal complex numbers have equal parts#

Suppose a+bi=c+dia + bi = c + di, where all four of aa, bb, cc, dd are real. Write r=acr = a - c and s=bds = b - d, both real. Subtracting the right side from the left turns the supposed equality into r+si=0r + si = 0, that is, r=sir = -si.

Square both sides. The left side becomes r2r^2, and the right side becomes (si)2=s2i2=s2(-si)^2 = s^2 i^2 = -s^2, so

r2=s2,that is,r2+s2=0.r^2 = -s^2, \qquad \text{that is,} \qquad r^2 + s^2 = 0.

Now rr and ss are real numbers, so r20r^2 \ge 0 and s20s^2 \ge 0. If either of them were nonzero, its square would be strictly positive and the sum r2+s2r^2 + s^2 would be strictly positive, not zero. So both squares are zero, which forces r=0r = 0 and s=0s = 0, meaning a=ca = c and b=db = d.

The converse direction is immediate: if a=ca = c and b=db = d, the two expressions are the same expression. So two complex numbers are equal exactly when their real parts match and their imaginary parts match.

This little theorem is what makes complex answers readable. When a computation ends at 23+2i23 + 2i, the real part 2323 and the imaginary part 22 are not one description among many; they are the only description. It also hands you a technique: an equation between complex numbers is really two equations between real numbers, one for the real parts and one for the imaginary parts. Several problems in the practice sets are solved by exactly that splitting.

Addition and subtraction combine like terms

Add two complex numbers by regrouping, which the commutative and associative laws permit, and then factoring ii out of the two imaginary terms, which the distributive law permits:

(a+bi)+(c+di)=(a+c)+(bi+di)=(a+c)+(b+d)i.(a + bi) + (c + di) = (a + c) + (bi + di) = (a + c) + (b + d)\,i.

Real parts add to real parts, and imaginary parts add to imaginary parts, precisely the way xx-terms and constant terms keep to themselves when you simplify (3+5x)+(2+4x)(3 + 5x) + (2 + 4x). Subtraction is the same computation after distributing the minus sign through both terms of the second number:

(a+bi)(c+di)=(ac)+(bd)i.(a + bi) - (c + di) = (a - c) + (b - d)\,i.

Notice what the right sides look like. Since a+ca + c and b+db + d are real numbers, each result is again (real)+(real)i(\text{real}) + (\text{real})\,i: adding and subtracting complex numbers can never produce anything but a complex number. Two operations down, two to go.

Worked example 1 Adding and subtracting

Compute (35i)+(7+2i)(3 - 5i) + (7 + 2i), then (4+3i)(68i)(4 + 3i) - (6 - 8i).

For the sum, collect the real parts and the imaginary parts separately:

(35i)+(7+2i)=(3+7)+(5+2)i=103i.(3 - 5i) + (7 + 2i) = (3 + 7) + (-5 + 2)\,i = 10 - 3i.

For the difference, first distribute the minus sign through the second number, turning (68i)-(6 - 8i) into 6+8i-6 + 8i, and only then collect:

(4+3i)(68i)=4+3i6+8i=(46)+(3+8)i=2+11i.(4 + 3i) - (6 - 8i) = 4 + 3i - 6 + 8i = (4 - 6) + (3 + 8)\,i = -2 + 11i.

The single most common slip in this entire lesson happens in that second computation: subtracting only the 66 and forgetting that the minus sign also flips the 8i-8i. Distribute first, then combine.

Check your understanding

Compute (53i)(2+4i)(5 - 3i) - (2 + 4i).

Answer choices

Multiplication is expansion plus one substitution

Multiply two complex numbers the way you multiply any two binomials: distribute every term of the first factor across every term of the second. Nothing about that step is new. The new step comes after, when an i2i^2 is sitting in the result and must be exchanged for 1-1:

(a+bi)(c+di)=ac+adi+bci+bdi2=(acbd)+(ad+bc)i.(a + bi)(c + di) = ac + ad\,i + bc\,i + bd\,i^2 = (ac - bd) + (ad + bc)\,i.

Read the middle expression carefully. The four products are exactly the four you would write for (a+bx)(c+dx)(a + bx)(c + dx). The only difference is the last one, where bdi2=bdbd\,i^2 = -bd abandons the imaginary column and joins the real part with its sign flipped. That is the whole novelty of complex multiplication. Do not memorize the boxed-looking formula (acbd)+(ad+bc)i(ac - bd) + (ad + bc)\,i; it is faster and safer to expand and substitute each time, letting the formula rebuild itself. And once more the result is (real)+(real)i(\text{real}) + (\text{real})\,i, so multiplication also stays inside the system.

Area model for (2 + 3i)(4 - 5i)A two-by-two grid of partial products. The i squared cell becomes real, so the product collects to 23 plus 2i.×4-5i23i8-10i12i-15i² = +15i² = -1, so the bottom-right cell lands back among the real numbersreal parts 8 + 15 = 23, imaginary parts -10i + 12i = 2i(2 + 3i)(4 - 5i) = 23 + 2i
An area model for the product of 2 + 3i and 4 - 5i. Each cell holds one of the four products from the expansion. Three cells are routine, but the bottom-right cell contains i squared, and the rule i squared = -1 folds it back into the real column. That returned +15 is why the real part of the answer is 23 rather than 8.

Worked example 2 Two products in full

Compute (2+3i)(45i)(2 + 3i)(4 - 5i), then (32i)2(3 - 2i)^2.

Expand the first product term by term, exactly as the area model above lays it out:

(2+3i)(45i)=810i+12i15i2.(2 + 3i)(4 - 5i) = 8 - 10i + 12i - 15i^2.

Now spend the rule. Since i2=1i^2 = -1, the last term is 15i2=+15-15i^2 = +15, a real number:

810i+12i+15=(8+15)+(10+12)i=23+2i.8 - 10i + 12i + 15 = (8 + 15) + (-10 + 12)\,i = 23 + 2i.

For the square, use the binomial pattern (xy)2=x22xy+y2(x - y)^2 = x^2 - 2xy + y^2 with x=3x = 3 and y=2iy = 2i, or simply expand:

(32i)2=912i+4i2=912i4=512i.(3 - 2i)^2 = 9 - 12i + 4i^2 = 9 - 12i - 4 = 5 - 12i.

In both computations the expansion was ordinary algebra; the only complex-number moment was trading each i2i^2 for 1-1 before collecting.

Check your understanding

Compute (3+i)(25i)(3 + i)(2 - 5i).

Answer choices

The conjugate makes products real

One special product deserves its own name, because it powers everything in the rest of the lesson. The conjugate of z=a+biz = a + bi is the number

z=abi,\overline{z} = a - bi,

the same number with the sign of its imaginary part flipped. So 4+7i=47i\overline{4 + 7i} = 4 - 7i, the conjugate of 35i-3 - 5i is 3+5i-3 + 5i, and 6i=6i\overline{6i} = -6i. Flipping the sign twice restores the original, so z=z\overline{\overline{z}} = z: conjugation is its own undo. A real number is its own conjugate, since r=r+0ir = r + 0i and flipping the sign of 00 changes nothing, and the matching-parts theorem gives the converse too: z=z\overline{z} = z demands b=b-b = b, so b=0b = 0. A complex number equals its own conjugate exactly when it is real.

Pair a number with its conjugate and both basic operations produce real answers. The sum is (a+bi)+(abi)=2a(a + bi) + (a - bi) = 2a, twice the real part, with the imaginary parts cancelling. The product is the important one.

The product of conjugates is real and never negative#

Expand the product of z=a+biz = a + bi and z=abi\overline{z} = a - bi like any other product:

(a+bi)(abi)=a2abi+abib2i2=a2+b2.(a + bi)(a - bi) = a^2 - ab\,i + ab\,i - b^2 i^2 = a^2 + b^2.

The two middle terms are opposites and cancel, and the last term flips sign because i2=1i^2 = -1. Compare this with the difference-of-squares pattern (x+y)(xy)=x2y2(x + y)(x - y) = x^2 - y^2 that you know from factoring: the pattern fires here too, giving a2(bi)2=a2b2i2a^2 - (bi)^2 = a^2 - b^2 i^2. That last i2i^2 turns the familiar difference of squares into a sum of squares. So

zz=a2+b2,z\,\overline{z} = a^2 + b^2,

a real number built from two squares of real numbers. Each square is at least 00, so zz0z\,\overline{z} \ge 0 always. Better still, the sum-of-squares argument from the equality proof applies word for word: a2+b2=0a^2 + b^2 = 0 forces a=0a = 0 and b=0b = 0, and conversely z=0z = 0 gives 02+02=00^2 + 0^2 = 0. So zzz\,\overline{z} is strictly positive for every complex number except z=0z = 0 itself.

Keep the shape of that result in mind: multiplying any complex number by its conjugate erases ii entirely and leaves the positive real number a2+b2a^2 + b^2. The next lesson will draw a+bia + bi as a point in a plane and read a2+b2a^2 + b^2 off the picture geometrically. For this lesson, what matters is that a2+b2a^2 + b^2 is real and, for z0z \ne 0, positive.

Conjugation has one more property worth proving now, because a famous fact about quadratics will fall out of it at the end of the lesson. Conjugation does not care whether you conjugate before or after doing arithmetic.

Conjugation respects sums and products#

Let z=a+biz = a + bi and w=c+diw = c + di be any two complex numbers, with a,b,c,da, b, c, d real.

For sums, compute both orders. Conjugating after adding gives z+w=(a+c)+(b+d)i=(a+c)(b+d)i\overline{z + w} = \overline{(a + c) + (b + d)\,i} = (a + c) - (b + d)\,i. Adding the conjugates gives z+w=(abi)+(cdi)=(a+c)(b+d)i\overline{z} + \overline{w} = (a - bi) + (c - di) = (a + c) - (b + d)\,i. The two agree, so z+w=z+w\overline{z + w} = \overline{z} + \overline{w}.

For products, the multiplication formula from earlier gives zw=(acbd)+(ad+bc)izw = (ac - bd) + (ad + bc)\,i, so conjugating after multiplying yields

zw=(acbd)(ad+bc)i.\overline{zw} = (ac - bd) - (ad + bc)\,i.

Multiplying the conjugates first instead,

zw=(abi)(cdi)=acadibci+bdi2=(acbd)(ad+bc)i,\overline{z}\,\overline{w} = (a - bi)(c - di) = ac - ad\,i - bc\,i + bd\,i^2 = (ac - bd) - (ad + bc)\,i,

the same number. So zw=zw\overline{zw} = \overline{z}\,\overline{w} as well. In words: conjugate first or compute first, the answer is identical, for addition and for multiplication. The answer is therefore identical for anything built out of those two operations, such as subtraction, squaring, and higher powers.

Worked example 3 Conjugate products at work

Compute (3+4i)(34i)(3 + 4i)(3 - 4i) and (1+2i)(12i)(1 + 2i)(1 - 2i) without a full expansion.

Each is a number times its own conjugate, so the theorem answers immediately with a2+b2a^2 + b^2:

(3+4i)(34i)=32+42=25,(1+2i)(12i)=12+22=5.(3 + 4i)(3 - 4i) = 3^2 + 4^2 = 25, \qquad (1 + 2i)(1 - 2i) = 1^2 + 2^2 = 5.

If you distrust the shortcut, expand the first one: 912i+12i16i2=9+16=259 - 12i + 12i - 16i^2 = 9 + 16 = 25. The cross terms cancel and the i2i^2 flips the last sign, exactly as the proof said they must. Notice both answers are positive real numbers; an ii can never survive this pairing.

Division is rationalizing in disguise

You have solved this problem before, in a different costume. When a denominator held a radical, you multiplied by a well-chosen form of 11 to clear it:

13+2=13+23232=3292=327.\frac{1}{3 + \sqrt{2}} = \frac{1}{3 + \sqrt{2}} \cdot \frac{3 - \sqrt{2}}{3 - \sqrt{2}} = \frac{3 - \sqrt{2}}{9 - 2} = \frac{3 - \sqrt{2}}{7}.

The conjugate 323 - \sqrt{2} was chosen because the difference-of-squares pattern makes the new denominator 32(2)2=73^2 - (\sqrt{2})^2 = 7, a rational number. Division of complex numbers is the same trick with the same name. To divide by c+dic + di, multiply numerator and denominator by its conjugate cdic - di, and the proof above guarantees the new denominator is the real number c2+d2c^2 + d^2:

a+bic+di=a+bic+dicdicdi=(a+bi)(cdi)c2+d2.\frac{a + bi}{c + di} = \frac{a + bi}{c + di} \cdot \frac{c - di}{c - di} = \frac{(a + bi)(c - di)}{c^2 + d^2}.

Multiplying by cdicdi\frac{c - di}{c - di} is multiplying by 11, so the value never changes; only its costume does. After this move, the numerator is one ordinary complex multiplication, and dividing a complex number by the positive real c2+d2c^2 + d^2 just divides each part, since p+qim=pm+qmi\frac{p + qi}{m} = \frac{p}{m} + \frac{q}{m}\,i.

Pause on when this is allowed. The move needs c2+d20c^2 + d^2 \ne 0, and the conjugate proof showed c2+d2=0c^2 + d^2 = 0 happens only for c+di=0c + di = 0. So the only forbidden division is division by zero, the same single exception the real numbers always had. Every nonzero complex number can divide, and in particular every nonzero z=c+diz = c + di has a reciprocal,

1c+di=cdic2+d2,\frac{1}{c + di} = \frac{c - di}{c^2 + d^2},

again of the form (real)+(real)i(\text{real}) + (\text{real})\,i. That settles the closure question asked at the start: all four operations take complex numbers to complex numbers. The system that began as one symbol and one rule supports the full arithmetic that the rational and real numbers enjoy, with nothing missing and nothing extra needed.

Worked example 4 A quotient, start to finish

Compute 3+2i14i\dfrac{3 + 2i}{1 - 4i}.

The denominator is 14i1 - 4i, so multiply numerator and denominator by its conjugate 1+4i1 + 4i:

3+2i14i1+4i1+4i=(3+2i)(1+4i)(14i)(1+4i).\frac{3 + 2i}{1 - 4i} \cdot \frac{1 + 4i}{1 + 4i} = \frac{(3 + 2i)(1 + 4i)}{(1 - 4i)(1 + 4i)}.

The denominator is a conjugate product, so it is 12+42=171^2 + 4^2 = 17 with no work. Expand the numerator and spend i2=1i^2 = -1:

(3+2i)(1+4i)=3+12i+2i+8i2=3+14i8=5+14i.(3 + 2i)(1 + 4i) = 3 + 12i + 2i + 8i^2 = 3 + 14i - 8 = -5 + 14i.

Now divide each part by 1717:

3+2i14i=5+14i17=517+1417i.\frac{3 + 2i}{1 - 4i} = \frac{-5 + 14i}{17} = -\frac{5}{17} + \frac{14}{17}\,i.

The answer is not pretty, but it is exact, and it has the required shape of a real part plus a real multiple of ii.

Worked example 5 Division undoes multiplication

Find the complex number zz satisfying (2i)z=4+3i(2 - i)\,z = 4 + 3i.

Division exists precisely to undo multiplication, so divide both sides by 2i2 - i, which is legal because 2i02 - i \ne 0:

z=4+3i2i=(4+3i)(2+i)(2i)(2+i)=8+4i+6i+3i222+12=5+10i5=1+2i.z = \frac{4 + 3i}{2 - i} = \frac{(4 + 3i)(2 + i)}{(2 - i)(2 + i)} = \frac{8 + 4i + 6i + 3i^2}{2^2 + 1^2} = \frac{5 + 10i}{5} = 1 + 2i.

Check it, because a check costs one multiplication:

(2i)(1+2i)=2+4ii2i2=2+3i+2=4+3i.(2 - i)(1 + 2i) = 2 + 4i - i - 2i^2 = 2 + 3i + 2 = 4 + 3i. \checkmark

Linear equations over the complex numbers solve exactly the way they did over the reals, because the algebraic laws are the same and the reciprocal always exists.

Check your understanding

Compute 2+i3i\dfrac{2 + i}{3 - i}.

Answer choices

The quadratics that had no solutions

Here is the payoff the chapter has been building toward. In the last chapter, the discriminant sorted every quadratic, and the verdict for Δ<0\Delta < 0 was blunt: no real solutions. Take z24z+5=0z^2 - 4z + 5 = 0, whose discriminant is Δ=(4)24(1)(5)=4\Delta = (-4)^2 - 4(1)(5) = -4. The quadratic formula asks for a number whose square is 4-4, and the reals have none; but the complex numbers do, because (2i)2=4i2=4(2i)^2 = 4i^2 = -4. Feeding 2i2i into the formula in place of the missing square root suggests the candidates

z=4±2i2=2±i.z = \frac{4 \pm 2i}{2} = 2 \pm i.

The formula was derived for real arithmetic, so treat this as a suggestion rather than a conclusion. Then let the arithmetic of this lesson give the honest verdict by direct substitution.

Worked example 6 Checking a complex root

Verify that z=2+iz = 2 + i satisfies z24z+5=0z^2 - 4z + 5 = 0.

Compute the square first, expanding and spending i2=1i^2 = -1:

z2=(2+i)2=4+4i+i2=3+4i.z^2 = (2 + i)^2 = 4 + 4i + i^2 = 3 + 4i.

Then assemble the whole left side, distributing 4-4 across 2+i2 + i:

z24z+5=(3+4i)(8+4i)+5=(38+5)+(44)i=0.z^2 - 4z + 5 = (3 + 4i) - (8 + 4i) + 5 = (3 - 8 + 5) + (4 - 4)\,i = 0.

Both the real part and the imaginary part collapse to 00, so 2+i2 + i really is a solution, certified by nothing but addition and multiplication.

What about the other candidate, 2i2 - i? It is the conjugate of the root just checked, and you do not need to repeat the computation to trust it. Conjugate both sides of the true statement z24z+5=0z^2 - 4z + 5 = 0. The right side is the real number 00, its own conjugate. The left side is built from zz by multiplication and addition with the real numbers 4-4 and 55. The proof above showed conjugation respects both operations while leaving real numbers fixed, so the conjugate of that left side is z24z+5\overline{z}^2 - 4\overline{z} + 5. The conjugated statement therefore reads z24z+5=0\overline{z}^2 - 4\overline{z} + 5 = 0: whenever a complex number solves this equation, its conjugate automatically solves it too. Notice also that the pair fits the sum and product of roots from the last chapter perfectly: (2+i)+(2i)=4(2 + i) + (2 - i) = 4 and (2+i)(2i)=22+12=5(2 + i)(2 - i) = 2^2 + 1^2 = 5, matching z24z+5z^2 - 4z + 5 term for term. The final lesson of this chapter turns these observations into the full story of quadratics with negative discriminants.

Common mistakes

Practice

Multiple Choice Questions (MCQ)

Progressively harder sets of questions. Each opens on its own page.

Free Response Questions (FRQ)

Longer questions in parts, to be worked out on paper. Progressive hints, the answer on its own so you can check yourself and try again, then the full worked solution, plus a rubric to mark your own work against.

Free response Work it out on paper 5 questions Start →
More practice (optional)

Extra sets, as hard as the Challenge set. Each one opens on its own page.

More resources (optional)

Other explanations of this lesson, if you want a second take.

A bit of history (Optional)

By the middle of the 1500s, square roots of negatives kept turning up inside perfectly normal problems. Nobody could ban them, since the algebra produced them without being asked. Nobody could use them with any confidence either. One writer had multiplied a pair of them in a single worked case. He set the answer aside as useless. What one such root times another meant in general, nobody had settled.

Rafael Bombelli was an engineer who drained marshes in northern Italy for a living. He wrote his algebra book while one of those projects was held up, and it appeared in 1572. What he offered was not so much a discovery as a decision. He treated the impossible root as an object like any other, and built it the first systematic table of signs.

He had no symbol for it, so he used words. A positive multiple of the root he called “plus of minus”, and a negative one “minus of minus”. Then came the rule that opened everything after it: plus of minus, times plus of minus, makes minus. Take the words away and that line is ii=1i \cdot i = -1, two centuries before the letter arrived.

The rest followed from ordinary algebra, and Bombelli used it on cubics that had defeated everyone. His key point was a humble one. Once you can multiply, you can expand, and this lesson is that point carried through. You spread the product out as you would any binomial, then trade every i2i^2 for 1-1.